- (database_info.c) Optimizations for database_info_build_query_enum() -
it would call strlcat at the very least 3 times, normally more than that.
Every strlcat call is at least one or more strlens. Reduce it all down
to manual string concatenation and one strlcat
- (database_info_build_query_enum) - always NULL terminates string so no need for NULL termination done by callee
* (menu_displaylist) use smaller strings for basename paths, remove one unnecessary string
in menu_displaylist.c function
- (disk_index_file.c) Avoid another trivial string concatenation by doing
it manually after fill_pathname_join_special
- (fill_pathname_join_delim) since fill_pathname_join_delim() now has a return value, use it for
string concatenation
- (platform_unix) - Use BUFSIZ as size for a buffer in system_property_get
- some other trivial strlcat usage replacements
- Don't read the entire archive file into memory on systems without HAVE_MMAP
- Support ZIP files between 2GB ~ 4GB on 32bit architectures
- Remove zip exclusive members from `file_archive_transfer`
- Add `zip_context_t` for zip like the already existing `sevenzip_context_t`
- Unify cleanups of zip and 7z in `file_archive_file_backend::archive_parse_file_free`
- Handle some sizes and offsets as int64_t (preparation for future Zip64 support)
- Handle compression methods in `zlib_stream_decompress_data_to_file_init` not in `file_archive_perform_mode`
- Return progress of 7z iterating with `file_archive_parse_file_progress`
Archive scans were broken for cores like Atari 5200 and Commodore 64 because archive contents were skipped if a core reported it supported compressed files. The problem is, it's a false assumption that just because a core supports compressed files that the content database contains CRCs for compressed files.
Instead of adding the contents of every compressed file to the scan list ahead of time, the patched logic now checks for the CRC of the archive itself as it always has but then adds the archive contents to the scan list if the archive itself is no match. This maintains the logic of adding only the archive if the core supports compressed files and the archive CRC matches but also allows for deeper content scans when there is no match.
The patch also removes `core_info_unsupported_content_path` as it was redundant with existing `core_info_database_supports_content_path` logic.
